Teleost Basophils Have IgM-Dependent and Dual Ig-Independent Systems Tomoyuki Odaka, Hiroaki Suetake, Tomoki Maeda and Toshiaki Miyadai This information is current as of September 28, 2021. J Immunol published online 7 March 2018 http://www.jimmunol.org/content/early/2018/03/07/jimmun ol.1701051 Downloaded from

Supplementary http://www.jimmunol.org/content/suppl/2018/03/07/jimmunol.170105 Material 1.DCSupplemental

Why The JI? Submit online. http://www.jimmunol.org/

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

*average by guest on September 28, 2021

Subscription Information about subscribing to The Journal of is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2018 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. Published March 7, 2018, doi:10.4049/jimmunol.1701051 The Journal of Immunology

Teleost Basophils Have IgM-Dependent and Dual Ig-Independent Degranulation Systems

Tomoyuki Odaka, Hiroaki Suetake, Tomoki Maeda, and Toshiaki Miyadai

Recently, mammalian basophils have been highlighted as having roles in and antiparasitic ; however, there is little information about the functions and evolutionary origin of basophils, because they are the least abundant leukocyte in most vertebrates. In this study, we characterized the teleost basophils that are abundant in the peripheral of fugu (Takifugu rubripes). Fugu basophils have two distinct granules: reddish-purple and dark violet ones. Teleost fish do not have IgG and IgE, but we found that fugu IgM bound on the surface of the basophils, and the cross-linked IgM induced degranulation of both types of granules. This indicates that teleost basophils can be activated in an Ab-dependent manner. Furthermore, papain induced the degranulation of the reddish-purple granules, which contain , and the released granules stimulated the migration of +

various leukocytes. In contrast, chitin elicited the degranulation of the dark violet granules, which resulted in CD4 –specific Downloaded from migration. Thus, fugu basophils control immune responses via two distinct Ab-independent mechanisms. In addition, fugu basophils endocytosed soluble Ag and expressed MHC class II and B7-H1/DC. These findings suggested that fugu basophils can interact with T cells as APCs. Thus, the Ab-dependent predates the emergence of IgG and IgE, and fish basophils exhibit different dynamics and features of degranulation to distinct stimuli compared with mammalian basophils. Some features of teleost basophils are more similar to those of mammalian mast cells than to those of mammalian basophils. The

Journal of Immunology, 2018, 200: 000–000. http://www.jimmunol.org/

asophils and mast cells are closely related that is absent in teleost fish (6). Therefore, we wonder whether teleost contribute to allergy and antiparasitic immunity in mammals fish have an Ig-dependent basophil-degranulation mechanism. B (1, 2). However, in contrast to well-characterized mast cells, Basophils also have an Ig-independent degranulation system the functions of basophils in the have not been fully (8–10). Some , such as chitin, are structural components understood, because basophils are the least abundant leukocyte in of parasites and fungi, and others are that are secreted by most vertebrate species. To understand the functions of basophils, parasites. However, little is known about the differences among some turtles and teleosts have been proposed to represent unique the cellular responses activated by these different types of aller-

animal models that have a natural abundance of basophils in their gens in basophils. by guest on September 28, 2021 circulating blood (3, 4). In this study, we focused on teleost basophils, To characterize the features of basophils in lower vertebrates, we which are more evolutionarily primitive than turtle basophils. This selected fugu (Takifugu rubripes), which is a teleost fish species study will yield important insights into the functions of basophils and whose genome has been sequenced (11). Fugu has a natural the difference between mast cells and basophils from an evolutionary abundance of basophils in peripheral blood (4), and it is easy to perspective. isolate the basophils because of their high density. In this study, Basophils release inflammatory mediators, such as histamine, by we characterized fish basophils and provide insights into the degranulation and induce inflammation (5). Mammalian basophils evolutionary history of vertebrate basophils. Teleost fish basophils have activation mechanisms mediated by IgE and IgG, which are have an Ig-dependent degranulation system and have two distinct exclusively present in mammalian species (6). The degranulation types of granules with distinct roles whose release is activated by of basophils or mast cells in nonmammalian species, such as different stimuli. reptiles and avians, has also been observed and is likely to be mediated by IgY, which is analogous to mammalian IgG and IgE Materials and Methods (3, 7). IgY emerged from amphibians, and an IgY/IgG/IgE analog Animals Fugu weighing between 300 and 800 g were bred at the Research Center for Marine Bioresources of Fukui Prefectural University. The fish were reared Faculty of Marine Science and Technology, Fukui Prefectural University, Obama, in a tank with seawater kept at 20˚C and were fed dry food pellets once a Fukui 917-0003, Japan day (Nippai). All experiments were done in accordance with the approved Received for publication July 21, 2017. Accepted for publication February 10, 2018. guidelines and regulations of Fukui Prefectural University. This work was supported in part by Yakult Grants for Research on Lactic Acid Bacteria (to T. Miyadai) and the Fukui Prefectural Government’s Grant-in-Aid for Cell preparation Collaboration Research Projects (to H.S.). Fish were anesthetized with 2-phenoxyethanol (Wako), and peripheral Address correspondence and reprint requests to Dr. Hiroaki Suetake, Fukui Prefec- blood was collected immediately using heparinized syringes and needles tural University, 1-1 Gakuen-cho, Obama, Fukui 917-0003, Japan. E-mail address: and centrifuged at 500 3 g for 20 min. After centrifugation, the peripheral [email protected] blood was separated into four layers: plasma, peripheral blood leukocytes The online version of this article contains supplemental material. (PBLks; including , , and nonhigh-density gran- Abbreviations used in this article: EIA, enzyme immunoassay; HDG, high-density ulocytes), erythrocytes, and high-density granulocytes (HDGs). HDGs pre- ; KLH, keyhole limpet hemocyanin; KLH-FITC, KLH coupled to FITC; cipitating at the tube bottom were collected and washed with PBS following MGG, May–Gru¨nwald–Giemsa; PBLk, peripheral blood leukocyte. the removal of plasma, PBLks, and erythrocytes. The PBLk layer excluding HDGs (described above) was further sepa- Copyright Ó 2018 by The American Association of Immunologists, Inc. 0022-1767/18/$35.00 rated from erythrocytes using Percoll (GE Healthcare). In brief, PBLks were

www.jimmunol.org/cgi/doi/10.4049/jimmunol.1701051 2 TELEOST BASOPHILS HAVE A NOVEL DEGRANULATION SYSTEM layered on 53% Percoll diluted with PBS and centrifuged at 2465 3 g for was detected (final concentration 0.016 EU/ml). It has long been estab- 35 min at 4˚C. After centrifugation, PBLks in the intermediate layer of lished that fish are resistant to the toxic effect of LPS (14), and there is no the Percoll gradient were collected and washed with PBS. ortholog in the fugu genome (15). Thus, it is reasonable that the effect For the endocytosis and assays, monocytes and erythro- of LPS in the chitin was excluded in our experiments. cytes were collected, respectively. Monocytes were collected as follows. The isolated PBLks were suspended in RPMI 1640 medium (Nissui Chemotaxis assay Pharmaceutical) and cultured at 22˚C for 1 h in tissue culture flasks Fugu HDGs were cultured for 4 h with chitin, papain, or heat-inactivated (Sumitomo Bakelite) coated with fugu plasma to enhance ad- papain (100 mg/ml). The culture medium was collected by centrifugation hesion to the flasks. Adherent monocytes were washed thoroughly several at 500 3 g for 10 min and was filtered using a 0.2-mm syringe filter times with PBS to remove nonadherent cells. Following the wash steps, the (Sartorius). This medium was used in the chemotaxis assay as the che- monocytes were stimulated with 50 mg/ml LPS from Escherichia coli O26 motactic solution. Chemotaxis was evaluated using a microchamber (Wako) at 22˚C for 24 h. The peripheral monocytes were collected for technique. Briefly, the chemotactic solution (600 ml per well) was placed further use. To isolate erythrocytes, after centrifugation of the peripheral in a 24-well plate (Sumitomo Bakelite), and Chemotaxicell chambers with blood, as described above, the erythrocyte layer was collected and washed a pore size of 3 mm (Kurabo) were filled with 200 ml of PBLk suspension with PBS. (4 3 106 cells per well) and incubated for 3 h at 22˚C. The filters were removed, and the migrated cells were stained with MGG for cell counting Cytology and morphological observation. The number of cells in five microscopic HDGs and PBLks (2 3 106 cells per milliliter) were suspended with PBS fields (3200 magnification) was counted using a phase-contrast micro- and analyzed using a FACSCalibur flow cytometer (BD Biosciences). One scope (Nikon). milliliter of HDG suspension (5 3 105 cells per milliliter) was smeared onto glass slides at 100 3 g for 5 min using a Cytospin centrifuge. For Detection of IgM on the basophil surface morphological observations, May-Gru¨nwald-Giemsa (MGG) was HDGs were incubated for 24 h at 20˚C with purified fugu IgM (1 mg/ml). Downloaded from used. The smears were fixed and stained for 2 min with May-Gru¨nwald Cells were washed with PBS, incubated for 30 min at 4˚C with an anti- solution and then stained for 2 min with 5 mM phosphate buffer (pH 8). fugu IgM mouse mAb (4H5) (16), washed again, and incubated for 30 min After the May-Gru¨nwald solution was removed, the specimens were at 4˚C with an FITC-labeled goat anti-mouse IgG (Jackson Immuno- stained for 7 min with Giemsa solution (Merck). To assay for meta- Research). After two washes with PBS, the cells were analyzed using a chromasia, cells were stained with toluidine blue (Chroma Gesellschaft), FACSCalibur flow cytometer. as described previously (12). In brief, the smears were fixed and stained for 5 min with 1% methanolic solutions of toluidine blue and rinsed in Induction of degranulation with IgM cross-linking aqueous solutions of hydrochloric acid (pH 3). http://www.jimmunol.org/ HDGs were incubated for 24 h at 20˚C with purified fugu IgM (1 mg/ml). RT-PCR and real-time PCR analyses Cells were washed with PBS and incubated for 1, 10, 20, 30, and 60 min at 20˚C with anti-fugu IgM mAb. The degranulated cells were counted at Total RNA was extracted from HDGs, PBLks, and kidney using an RNeasy each time point using a Bu¨rker-Tu¨rk chamber, and they were also smeared Mini Kit and an RNase-Free DNase Set (both from QIAGEN). Total RNA 2 onto glass slides and observed by MGG staining. (50 ng) was reverse-transcribed into cDNA using SuperScript III RNase H In parallel, HDGs cultured with purified fugu IgM, as described above, Reverse Transcriptase (Invitrogen). First-strand cDNA was amplified using were incubated for 10 min at 20˚C with anti-fugu IgM mAb after washing gene-specific primer sets (Supplemental Table I) and KAPA Taq EXtra with PBS. The cells were washed with PBS and incubated for 30 min at HotStart ReadyMix with dye (Kapa Biosystems). The PCR conditions were 4˚C with FITC-labeled goat anti-mouse IgG. After two washes with PBS, as follows: 2 min initial denaturation at 95˚C and then 30–35 cycles with the cells were smeared onto glass slides and observed using a fluorescent by guest on September 28, 2021 30 s denaturation at 95˚C, 30 s annealing at 60˚C, and 30 s amplification at microscope to ascertain whether the degranulated cells were IgM+. 72˚C. After 5 min of terminal amplification at 72˚C, the samples were run on Furthermore, we investigated the morphology of IgM+ cells that were a 2% agarose gel with ethidium bromide for visualization. purified by MACS. In brief, HDGs were incubated for 24 h at 20˚C with To investigate the expression of markers in migrated cells purified fugu IgM (1 mg/ml). Then, cells were incubated for 10 min at that were induced by chitin stimulated basophils supernatant, the migrated 20˚C with anti-fugu IgM mAb, washed with PBS, and incubated for 30 min at leukocytes were collected. Total RNA was extracted by the same method as 4˚C with anti-mouse IgG MicroBeads (Miltenyi Biotec). After additional wash described above. Thirty nanograms of total RNA was reverse-transcribed steps, IgM+ cells were separated using a SuperMACS cell separator (Miltenyi into cDNA using PrimeScript II Reverse Transcriptase (Takara). Quanti- Biotec), following the instructions provided by the manufacturer. The mor- tative PCR was performed using GeneAce SYBR qPCR Mix a No ROX phology of the purified cells was observed by MGG staining. (Nippon Gene) and CFX96 Real-Time System (Bio-Rad) in the DD cycle threshold (Ct) method. Gene-specific primer sets used for real-time PCR Endocytosis and phagocytosis assay were shown in Supplemental Table I. Each sample was amplified in trip- 6 licate. The gene expression levels were normalized by ef1a. For RT-PCR HDGs, RBCs, and (2 3 10 cells per well) were incubated for and real-time PCR, at least three individuals were used for each target 4 h at 22˚C with 70 mg/ml keyhole limpet hemocyanin (KLH) coupled to gene. FITC (KLH-FITC; Wako) in a six-well plate (Sumitomo Bakelite). After the incubation, cells were treated with 0.1% trypsin-PBS at 22˚C for Detection of histamine in fugu HDGs 10 min, washed, and analyzed using a FACSCalibur flow cytometer. In addition, the cells were smeared onto glass slides and observed using a For the detection of histamine using an enzyme immunoassay (EIA), HDGs fluorescent microscope. and PBLks (1 3 106 cells) were lysed with PBS containing 0.2% Triton X- HDGs or macrophages (2 3 106 cells per well) were cocultured with 100. The cell-free supernatants were collected, and the concentration of 1-mm latex beads (2 3 107 particles per well) for 4 h at 22˚C in a six-well histamine in each sample was measured using a histamine EIA (Oxford plate. The cells were washed with PBS and analyzed using a FACSCalibur Biomedical Research), in accordance with the instructions provided by the flow cytometer. manufacturer. For histamine detection in cytoplasmic granules, the cells were stained Expression analysis of costimulatory molecules with o-phthalaldehyde (Nacalai Tesque), according to the method de- HDGs (2 3 106 cells per well) were incubated with or without KLH scribed by Juhlin and Shelley (13). In brief, the specimens were incubated (70 mg/ml) in a six-well plate for 4 h at 22˚C. Subsequently, cells were for 3 min with 1% o-phthalaldehyde in xylene and then mounted using collected and reacted for 1 h at 4˚C with normal rat serum, anti–B7-H1/ tetrahydrofurfuryl alcohol. Within 3–10 min, the specimens were viewed under a fluorescent microscope (Olympus) equipped with a digital CCD DC, or anti–B7-H3 and anti–B7-H4 antiserum (dilution 1:500) (17), camera (ORCA-3CCD; Hamamatsu Photonics). washed with PBS, and incubated for 30 min at 4˚C with FITC-goat anti-rat For the histamine-release assay, HDGs (4 3 106 cells per well) were IgG (dilution 1:50; Invitrogen). After the wash steps, the basophils were analyzed using a FACSCalibur flow cytometer. stimulated for 4 h at 22˚C with 100 mg/ml chitin (40–70 mm), papain (Boehringer Mannheim), or heat-inactivated papain in a 24-well plate. Statistical analysis Then, the cell culture supernatants were collected, and histamine con- centration in the supernatants was quantified by EIA. A ToxinSensor All numerical data are presented as the mean 6 SEM. Data were analyzed chromogenic LAL Endotoxin Assay Kit (GenScript) was used to measure using repeated-measures ANOVA, followed by the Tukey method. Dif- the concentration of LPS in the chitin. No LPS contamination of the chitin ferences were considered significant at p , 0.05 or p , 0.01. The Journal of Immunology 3

Results histamine per 106 cells (34). This indicates that histamine release from HDGs in fugu peripheral blood are basophils the reddish-purple granules in the fugu basophils is induced by cys- teine stimulation. Next, we investigated whether the fugu To characterize fish basophils, we collected HDGs, which were basophils induce the migration of PBLks. For a chemotaxis assay, designated as basophils in a previous study (18), from fugu peripheral conditioned medium was obtained from the basophils stimulated with blood. After centrifugation of the blood, three cell layers appeared. papain. Various leukocytes, including lymphocytes, monocytes, and From the least to most dense, these layers represented PBLks, , migrated toward the conditioned medium of the basophils erythrocytes, and HDGs. We further analyzed the scatter profile of stimulated with papain (Fig. 2C, Supplemental Fig. 1D, 1E). fugu HDGs by flow cytometry, which demonstrated that they were a In contrast, the dark violet granules degranulated upon stimulation homogeneous cell population with an FSCloSSChi scatter profile, with chitin (Fig. 2A). These results strongly suggested that the two different from that of PBLks (Fig. 1A). This result suggested that types of granules in fugu basophils have different induction mecha- HDGs are highly granulated cells. The HDG population made up nisms for their degranulation. The amount of histamine released from 1–14% (mean 4.12 6 0.68%, n = 25) of total PBLks in fugu. basophils stimulated with chitin was comparable to that released from To observe the morphological and staining features of HDGs, the unstimulated control (Fig. 2B), which indicated that histamine is cells were stained with MGG stain. We observed that HDGs not present in the dark violet granules. Therefore, we examined contained two types of granules: small reddish-purple whether leukocyte migration was induced by the degranulation of the granules and large dark violet ones (Fig. 1B). In addition, they dark violet granules. The conditioned medium from cells stimulated had an eccentric round nucleus, similar to that of basophils from with chitin induced significant chemotaxis of lymphocytes (Fig. 2D). amphibians to birds and that of mammalian mast cells (Fig. 1B) To characterize the migrated lymphocytes, lymphocyte marker ex- Downloaded from (19–22). To characterize the cytochemistry of HDGs, we stained pression was analyzed by real-time PCR. The migrated lymphocytes them with toluidine blue. Our results showed that they exhibited exclusively expressed tcra and cd4 (Fig. 2E). Although ighm and tcrd metachromasia (Fig. 1B), which is a feature of basophils and mast were also expressed, a significant difference was not detected among cells. This result supports that HDGs are basophils. thesamples(Fig.2E).Incontrast,cd8a (CD8a for CD8+ T cells), To obtain further evidence that HDGs are basophils, we next prf1,andight (IgT H chain for IgT+ B cells) were not detected (data investigated gene expression. RT-PCR demonstrated that HDGs

not shown). Thus, we demonstrated that the migrated lymphocytes http://www.jimmunol.org/ expressed cebpa (encoding C/EBPa) but not kit (encoding c-kit) were CD4+ T cells. Next, to identify which type of CD4+ T cells (Fig. 1C). In mammals, C/EBPa is expressed in basophils but not migrated, the expression of some T cell subset markers was exam- in mast cells (23, 24), and c-kit shows the opposite expression ined. The markers tbx21 (T-bet for Th1), gata3 (GATA-3 for Th2), pattern in mice (25). In addition, RT-PCR showed that no other and foxp3 (Foxp3 for regulatory T cells) were not expressed by the leukocyte markers were detected in HDGs (Fig. 1C). We have the migrated cells (data not shown), suggesting that the migrated T cells additional evidence that fugu HDGs were circulating in the pe- were not effector T cells. For further characterization, the naive T cell ripheral blood. In mammals, basophils and immature mast cells marker ccr7 was analyzed and found to be expressed by the T cells are present in the circulating peripheral blood (26, 27). Unlike (Fig. 2E). In addition, fugu basophils constitutively expressed il16,

fugu HDGs, human progenitors lack secretory granules by guest on September 28, 2021 which is a chemoattractant of mammalian CD4+ cells (Fig. 2F). (28). Taken together, these results indicate that HDGs found in These results suggested that chitin-stimulated basophils release an fugu peripheral blood are fish basophils. attractant of naive CD4+ T cells. Thus, we demonstrated that fish basophils respond to allergens in an Ab-independent manner. Two distinct types of granules in fugu basophils are degranulated by different stimuli in an Ab-independent manner IgM-dependent degranulation occurs in fugu basophils We characterized the morphology and gene-expression profiles of It has not been revealed whether fish basophils are involved in Ab- fugu basophils. There is little prior information regarding the dependent immune responses, similar to mammalian basophils. function of fish basophils. The primary of mammalian basophils is to Mammalian basophils bind IgG and IgE (35); however, IgG and release various factors by degranulation, and the released factors IgE are not present in fish (6). Thus, key questions remain to be induce a variety of physiological effects. Basophils are activated via addressed. For example, how do the basophils recognize Ags? Ig-dependent and Ig-independent mechanisms (29–32). To under- Are teleost basophils activated via Ag-specific recognition by Ig stand the immunological basis for the Ab-independent degranula- molecules? Therefore, we investigated whether fugu basophils tion of fugu basophils, we investigated whether they degranulated have an Ab-dependent response. In this study, we focused on upon stimulation with papain and chitin as representative stimuli IgM, which is the major Ab in fish blood. connected with the accumulation and activation of basophils during First, we analyzed whether IgM can bind on the basophil surface. the immunological response to parasites and allergic immunity, We found that fugu basophils could bind to IgM, but the percentage respectively (31, 33). of IgM+ basophils varied among individual fish (Fig. 3A), and Papain stimulation induced the degranulation of the reddish- most of the tested fish showed a very low binding rate (,1% of purple granules only, but heat-inactivated papain did not induce cells). These observations suggested that IgM allows teleost ba- degranulation (Fig. 2A). This result suggested that the enzymatic sophils to respond to Ags in an Ag-specific manner. Hence, we activity of papain is important to stimulate the degranulation of the attempted to increase IgM binding on the basophil surface by fugu basophils. In mammals, basophils release histamine in re- incubating the cells with a high concentration of IgM. The per- sponse to various stimuli (5). In this study, we assayed whether centage of IgM binding on fugu basophils increased (Fig. 3B, 3C), histamine release from the reddish-purple granules was induced by which indicates that the response of fugu basophils to IgM is papain stimulation. An EIA demonstrated that the concentration of likely to be mediated by an inducible . histamine was significantly higher in the supernatant of the papain- Next, we examined whether the degranulation of basophils stimulated basophils than in the nonstimulated control (Fig. 2B). In occurs upon the cross-linking of surface IgM by an anti-fugu IgM addition, fugu basophils contained 9.2 6 0.6 ng histamine per 106 mAb. IgM cross-linking on fugu basophils induced degranulation cells (Supplemental Fig. 1A). This concentration is much lower (Fig. 3D). In contrast, this was not observed in basophils cultured than that in human basophils, which was reported as 2.4 6 0.6 mg with normal mouse serum, normal mouse serum with fugu IgM, or 4 TELEOST BASOPHILS HAVE A NOVEL DEGRANULATION SYSTEM Downloaded from

FIGURE 1. Fugu HDGs are basophils. (A) scatter profiles for fugu HDGs and PBLks. Dead cells were stained by propidium iodide and excluded from the data. (B) Morphological and histochemical appearance of HDGs after MGG and http://www.jimmunol.org/ toluidine blue (TB) staining. Scale bar, 10 mm. Arrows and arrowheads indicate small reddish-pur- ple granules and large dark violet ones, respectively. (C) Expression analysis of leukocyte marker genes in fugu HDGs, kidney, and PBLks. All data are representative results from three individuals. actb, b-actin for internal control; cd3e, CD3ε for T cell; cd83, CD83 for ; cebpa, C/EBP-a for basophil; ighm, IgM H chain for ; kit, c-kit for by guest on September 28, 2021 mast cell; lcp1, L-plastin for monocyte; M, molec- ular size marker; mpo, for neu- trophil; NTC no-template control; prf1, perforin for cytotoxic cell; tcra, TCRa for abT cell.

anti-IgM mAb alone (Supplemental Fig. 2). In mammals, the reached a maximum between 1 and 10 min, and the maximum degranulation caused by the cross-linking of IgE is completed percentage was equal to the proportion of IgM+ basophils among within a few minutes after stimulation (36). We examined the the cells (Fig. 3E). Because fugu basophils quickly degranulated temporal dynamics of the degranulation of fugu basophils after after IgM cross-linking, the response resembled anaphylactic the cross-linking of IgM. The percentage of degranulated cells degranulation by IgE cross-linking in mammalian basophils. To The Journal of Immunology 5

FIGURE 2. Fugu basophils exhibit distinct degranulation patterns to different Ig-inde- pendent stimuli. (A) Morphology of fugu ba- sophils that were stimulated by papain, chitin,

or heat-inactivated (HI) papain. Cells were Downloaded from stained with MGG stain. Arrows and arrow- heads indicate small reddish-purple granules and large dark violet ones, respectively. Scale bars, 10 mm. (B) Histamine-release assay in fugu basophils. Cells were stimulated with papain, chitin, and heat-inactivated (HI) pa- pain, and the concentration of histamine in the http://www.jimmunol.org/ culture supernatants was measured by EIA. (C) Chemotaxis of PBLks to the culture su- pernatant of papain-stimulated fugu basophils. Values were normalized to that of the non- stimulated control. (D) Chemotaxis of PBLks to the culture supernatant of chitin-stimulated fugu basophils. (E) Gene expression of fugu lymphocyte markers in migrated PBLks. The expression level of each gene was normalized by guest on September 28, 2021 by ef1a. The expression level in the cells migrated to the supernatant of chitin-stimu- lated basophils was set as 1. (F) Expression of il16 in fugu basophils. All data are represen- tative results from three individuals; error bars show the SEM of three independent experi- ments in parallel. *p , 0.05, **p , 0.01. cd4, CD4 for CD4+ T cell; ccr7, CCR7 for naive T cell; M, molecular size marker; ND, not de- tected; NTC, no-template control; tcrd, TCRd for gd T cell.

confirm whether the degranulated basophils were IgM+,we degranulation was observed in IgM2 cells. To ascertain whether observed the morphology of IgM+ basophils that were treated the degranulation is an artifact caused by the use of MACS, we with anti-IgM mAb, and the cells were sorted by MACS. Most performed immunostaining of IgM on the surface of the basophils. of the IgM+ basophils degranulated (Fig. 3F). In contrast, no The degranulated cells were IgM+ basophils (Fig. 3G). These 6 TELEOST BASOPHILS HAVE A NOVEL DEGRANULATION SYSTEM

FIGURE 3. IgM cross-linking–induced degranula- tion of fugu basophils. (A) IgM binds on the surface of fugu basophils. Basophils were stained with an anti- Downloaded from fugu IgM mAb. Gates indicate IgM+ basophils. (B) IgM+ basophils increased in coculture with a high concentration of IgM (1 mg/ml). Basophils were stained with an anti-fugu IgM Ab. IgM+ basophils were gated. (C) The percentage of IgM+ basophils increased significantly after IgM treatment. (D) Fugu http://www.jimmunol.org/ basophils degranulated by IgM cross-linking with anti-fugu IgM mAb. Cells were stained with MGG stain. Scale bar, 10 mm. (E) Time course of the change in the proportion of degranulated basophils. The ratio of degranulated cells/total basophils is shown for each time point after IgM cross-linking. (F) Morphology of magnetically sorted IgM+ or IgM2 basophils. Scale bars, 20 mm. (G) Immunofluorescence of basophils stained with anti-fugu IgM. The image showing FITC emission was colorized using the HCImage program by guest on September 28, 2021 supplied with the ORCA-3CCD microscope camera (Hamamatsu Photonics). The arrow shows a degra- nulated basophil. Scale bars, 10 mm. The data are representative results from three individuals. Dead cells were stained by propidium iodide and excluded from flow cytometry data. Error bars show the SEM of three independent experiments in parallel. **p , 0.01. DIC, differential interference contrast.

results indicated that fish basophils have an Ab-dependent de- APCs. Recent studies have revealed that mammalian basophils are granulation system. APCs that preferentially induce Th2 cells (37–39). However, mouse basophils are not competent to acquire Ags (40) and do not Fugu basophils have the capacity to function as Th2-inducing APCs express MHC class II genes. Nevertheless, they can present Ags Because fugu basophils induced the migration of CD4+ T cells by acquiring an MHC–peptide complex by trogocytosis from (Fig. 2D, 2E), we hypothesized that fugu basophils function as dendritic cells (41). The Journal of Immunology 7 Downloaded from http://www.jimmunol.org/ by guest on September 28, 2021

FIGURE 4. Fugu basophils endocytose soluble Ag and upregulate the expression of B7-H1/DC. (A) Endocytosis of KLH-FITC in fugu basophils, monocytes, and erythrocytes. Cells were treated with trypsin to remove the nonspecific binding of KLH-FITC on the cell surface. The black line and gray-shaded graphs indicate trypsinized and nontrypsinized cells, respectively. (B) Microscopic image of KLH-endocytosed basophils. This image is a merge of the differential interference contrast and FITC emission images. The image showing FITC emission was colorized by the HCImage program supplied with the ORCA-3CCD camera. Arrows indicate KLH-endocytosed basophils. Scale bar, 10 mm. (C)Phagocytosisof1-mm-diameter fluorescent latex beads by fugu basophils and monocytes. (D) Expression of MHC class II gene in fugu basophils, with or without KLH. (E)FACS (Figure legend continues) 8 TELEOST BASOPHILS HAVE A NOVEL DEGRANULATION SYSTEM

To clarify whether fugu basophils function as APCs, we first undergo Ig-dependent degranulation by IgM, IgD, and IgT. In this investigated whether they could take up soluble and particulate study, we demonstrated that fugu basophils underwent degranu- Ags. Interestingly, 11.9% of fugu basophils endocytosed a soluble lation in an IgM-mediated manner (Fig. 3). This finding indicated Ag: KLH-FITC (Fig. 4A). In addition, fluorescent mi- that Ab-dependent degranulation emerged at an early stage of croscopy demonstrated that FITC+ endosomes were formed in vertebrate evolution. In addition, we also showed that non-IgG and ∼11% of basophils (Fig. 4B). In contrast, compared with mono- non-IgE Abs can induce degranulation in basophils. In teleosts, it cytes (phagocytic rate 13.2%), basophils were far less efficient in has been reported that channel catfish granulocytes were also their phagocytosis of fluorescent latex beads (phagocytic rate armed with IgD (47–49), but the function of teleost IgD and IgT 2.57%, Fig. 4C). Thus, fish basophils are competent to preferen- on mast cells and basophils has not been fully investigated. It tially endocytose a soluble protein Ag. could be useful to investigate the possibility of novel Ig-dependent We demonstrated that fugu basophils endocytosed a soluble Ag, degranulation in these granulocytes. Fish IgM can increase the which implies that teleost basophils function as APCs. Therefore, we percentage of IgM+ basophils (Fig. 3B). The expression of FcεRI examined the expression and regulation of MHC class II and some on the cell surface correlates with the concentration of IgE in costimulatory factors in fugu basophils. In contrast to mammalian mammalian basophils (50–52). The findings of this study sug- basophils, fugu basophils constitutively expressed the MHC class IIa gested that there is a similar mechanism involving an Fc receptor gene (Fig. 4D). Flow cytometry demonstrated that fugu basophils for IgM and its ligand on fish basophils. expressed B7-H1/DC on their cell surface. Furthermore, the pro- Histamine was originally thought to be absent in fish leukocytes portion of B7-H1/DC+ cells increased after the endocytosis of KLH (53); however, Mulero et al. (54) reported that some fish mast cells

(Fig. 4E). Fugu B7-H3 and B7-H4 were not expressed on basophils, stored histamine. Reptile, avian, and mammalian basophils con- Downloaded from with or without KLH (Fig. 4E). These results suggested that fugu tain histamine in their granules (55, 56); however, it has been basophils have the potential to function as APCs. believed that amphibian and fish basophils do not contain hista- Mouse basophils are critical for Th2 differentiation owing to mine (53, 55). In this study, we demonstrated that teleost basophils their production of IL-4 production, as well as their presentation of store histamine in their granules (Fig. 2, Supplemental Fig. 1A, Ags (36–39). Therefore, we hypothesized that fugu basophils are 1B). This indicated that the production of histamine in basophils

also involved in Th2 differentiation via secretion of Th2 . has already been established in fish, which is also supported by http://www.jimmunol.org/ Fish have two orthologs of the il4 and il13 genes: il4/13a and il4/ several previous reports. Amphibian and fish mast cells also 13b (42). In mammals, il4 and il13 are derived from the tandem contain histamine (54, 57). We detected the expression of the duplication of an ancestral gene, whereas il4/13a and il4/13b gene encoding , which is an essential originated from a whole-genome duplication in teleost species. enzyme for the production of histamine (58), in fugu basophils These cytokines enhance the expressions of Th2 and (Supplemental Fig. 1C). In addition, ascidian basophil/mast cell– transcription factors, as well as B cell proliferation and Ab pro- like granular cells and test cells contain histamine in their granules duction, in fish (43, 44). In this study, we examined the expression (59, 60). Thus, histamine in granulocytes has an ancient origin. of these cytokines in fugu basophils. RT-PCR analysis showed that We speculate that the common ancestor of ascidians and verte- fugu basophils expressed il4/13a and il4/13b (Fig. 4F). These brates had histamine in basophil/mast cell–like granular cells. by guest on September 28, 2021 results suggested that fugu basophils are related to the Th2-like Histamine was released in association with degranulation, but the regulatory system as IL-4/13–producing cells. histamine content in fugu basophils (9.2 6 0.6 ng per 106 cells) is much lower than that in human basophils (2.4 6 0.6 mg per 106 Discussion cells) (28). In invertebrates, although the histamine content in test To understand the functions of basophils in nonmammalian ver- cells is comparable to that in fish, histamine plays a role in defense tebrates, we focused on fugu, a teleost species that has abundant mechanisms (58, 61, 62). In addition, Mulero et al. (54) reported basophils in peripheral blood (4). We could easily isolate fugu that gilthead sea bream could respond to histamine. Their obser- basophils based on their high density, which allowed us to char- vations suggested that fish have functional histamine receptors. acterize their function. We demonstrated that fugu basophils have We concluded that fugu basophils have the potential to be in- Ig-dependent and Ig-independent degranulation systems. Fugu volved in inflammation via the release of histamine. basophils have two types of granules, and different stimuli, such Papain stimulation induced histamine release from fugu basophils as IgM cross-linking, papain, and chitin, produced distinct de- (Fig. 2B). In contrast, papain stimulation induced IL-4 production, granulation patterns. These findings suggest that fugu basophils but not histamine release, in mouse basophils (31, 63). This sug- have different degranulation pathways. Thus, teleost basophils gests that fish basophils are directly activated by papain via a have already established Ig-dependent and Ig-independent im- pathway that is distinct from that in mammalian basophils. Inter- mune responses. estingly, human mast cells recognized papain directly via an un- Mammalian basophils have IgE- and IgG-mediated immune known receptor and underwent degranulation (64). It seems that fish responses (35, 45), and turtle basophils can bind to IgY, which is basophils have a pathway to release histamine via papain activation known as an ancient type of IgG and IgE (3). Teleost species do that is similar to the one in mammalian mast cells. not have IgE, IgG, or IgY; they only have IgM, IgD, and IgT. In We demonstrated that chitin is directly recognized by fugu zebrafish, it was suggested that the mast cells possess an FcεRI- basophils and can induce the degranulation of their dark violet like receptor, and cross-linking of the receptors by mouse IgE granules. This suggests that fugu basophils have multiple pathways and Ag induced degranulation (46). However, there has been for degranulation. In contrast to the reddish-purple granules, the a question of whether the basophils and mast cells of teleosts dark violet granules of fugu basophils did not contain histamine or

staining of B7 molecules on fugu basophils, incubated with or without KLH. The gates indicate cells positively stained for B7 molecules. (F) Expression analysis of il4/il13 in fugu basophils. All data are representative of at least three independent experiments. Dead cells were stained by propidium iodide and excluded from flow cytometry data. NTC, no-template control. The Journal of Immunology 9 induce CD4+ T cell migration (Fig. 2). These findings suggest that 3. Mead, K. F., M. Borysenko, and S. R. Findlay. 1983. Naturally abundant ba- the reddish-purple and dark violet granules have different func- sophils in the snapping turtle, Chelydra serpentina, possess cytophilic surface with reaginic function. J. Immunol. 130: 334–340. tions and that fugu basophils secrete a T cell–specific chemo- 4. Suzuki, Y. 1992. Basophil migration in acute inflammation in the puffer fish. attractant when stimulated with chitin. Nippon Suisan Gakkaishi 58: 2005–2007. 5. Karasuyama, H., K. Mukai, K. Obata, Y. Tsujimura, and T. Wada. 2011. Non- Fugu basophils constitutively expressed il16 mRNA, similarly to redundant roles of basophils in immunity. Annu. Rev. Immunol. 29: 45–69. mammalian mast cells (Fig. 2F). In mammals, basophils also par- 6. Pettinello, R., and H. Dooley. 2014. The immunoglobulins of cold-blooded ticipate in CD4+ T cell migration (65). IL-16 is a chemoattractant vertebrates. Biomolecules 4: 1045–1069. + 7. Caldwell, D. J., H. D. Danforth, B. C. Morris, K. A. Ameiss, and A. P. McElroy. foravarietyofCD4 T cells (66), and IL-16 produced by human 2004. Participation of the intestinal epithelium and mast cells in local mucosal + mast cells specifically induces CD4 T cell migration (67). There- immune responses in commercial poultry. Poult. Sci. 83: 591–599. fore, we assumed that fugu basophils can attract CD4+ T cells by 8. Kurimoto, Y., A. L. de Weck, and C. A. Dahinden. 1989. 3-dependent mediator release in basophils triggered by C5a. J. Exp. Med. 170: 467–479. secreting IL-16, similar to mammalian mast cells. Furthermore, we 9. Bischoff, S. C., A. L. de Weck, and C. A. Dahinden. 1990. and demonstrated that fugu basophils incorporated a soluble Ag and granulocyte/-colony-stimulating factor render human basophils re- expressed the MHC class IIa gene and B7-H1/DC (Fig. 4). Our sponsive to low concentrations of complement component C3a. Proc. Natl. Acad. Sci. USA 87: 6813–6817. results suggested that fish basophils have the potential to control 10. Bischoff, S. C., T. Brunner, A. L. De Weck, and C. A. Dahinden. 1990. Inter- T cell activities as a type of APC. Thus, it is suggested that teleost leukin 5 modifies histamine release and generation by human ba- basophils participate in innate immunity, as well as adaptive im- sophils in response to diverse agonists. J. Exp. Med. 172: 1577–1582. 11. Aparicio, S., J. Chapman, E. Stupka, N. Putnam, J. M. Chia, P. Dehal, munity. Mouse basophils were not competent to acquire Ags (40) A. Christoffels, S. Rash, S. Hoon, A. Smit, et al. 2002. Whole-genome shotgun and did not express MHC class II genes, but they could present Ags assembly and analysis of the genome of Fugu rubripes. Science 297: 1301–1310. by acquiring an MHC–peptide complex by trogocytosis from den- 12. Suzuki, Y. 1986. Cytochemistry of basophil granulocyte in carp and puffer. Bull. Downloaded from Jap. Soc. Sci. Fish. 52: 1895–1899. dritic cells (50). In contrast, mammalian mast cells have phagocytic 13. Juhlin, L., and W. B. Shelley. 1966. Detection of histamine by a new fluorescent capabilities (68) and function as APCs (69). Our study demon- omicron-phthalaldehyde stain. J. Histochem. Cytochem. 14: 525–528. strated that fugu basophils have the potential to act as APCs; this 14. Berczi, I., L. Berto´k, and T. Bereznai. 1966. Comparative studies on the toxicity of Escherichia coli lipopolysaccharide endotoxin in various animal species. Can. feature is shared with mammalian mast cells. However, the inter- J. Microbiol. 12: 1070–1071. action site between basophils and CD4+ T cells remains unknown. 15. Oshiumi, H., T. Tsujita, K. Shida, M. Matsumoto, K. Ikeo, and T. Seya. 2003. Prediction of the prototype of the human Toll-like receptor gene family from the B7-H1/DC was induced in fugu basophils cocultured with Ag http://www.jimmunol.org/ pufferfish, Fugu rubripes, genome. Immunogenetics 54: 791–800. (Fig. 4E). In our previous study, fugu B7-H1/DC suppressed T cell 16. Miyadai, T., M. Ootani, D. Tahara, M. Aoki, and K. Saitoh. 2004. Monoclonal proliferation and stimulated IL-10 expression (23). In mammals, recognising serum immunoglobulins and surface immunoglobulin- APCs highly expressing B7-DC develop CD4+ Tcellmemory(70). positive cells of puffer fish, torafugu (Takifugu rubripes). Fish Shellfish Immu- nol. 17: 211–222. In addition, mouse B7-H1 induces type 1 regulatory T cells, which 17. Sugamata, R., H. Suetake, K. Kikuchi, and Y. Suzuki. 2009. Teleost B7 produce suppressive cytokines, such as IL-10 and TGF-b (71), and expressed on monocytes regulates T cell responses. J. Immunol. 182: 6799–6806. + 18. Suzuki, Y. 1988. Isolation of heavy basophils in carp peripheral blood by Percoll CD4 regulatory T cells, which contribute to memory formation density gradient centrifugation. Nippon Suisan Gakkaishi 54: 1257. through IL-10 production (72). Interestingly, mammalian basophils 19. Turner, R. J. 1988. Amphibians. In Vertebrate Blood Cells. A. F. Rowley, and enhance immunological memory (73). Therefore, we speculate that N. A. Ratcliffe, eds. Cambridge University Press, New York, p. 129–209.

+ 20. Sypek, J., and M. Borysenko. 1988. Reptiles. In Vertebrate Blood Cells. A. F. by guest on September 28, 2021 B7-H1/DC basophils might also enhance immunological memory. Rowley, and N. A. Ratcliffe, eds. Cambridge University Press, New York, In our study, fugu basophils share common features with p. 210–256. mammalian mast cells. Generally, it is known that basophils share a 21. Maxwell, M. H. 1973. Comparison of heterophil and basophil ultrastructure in six species of domestic birds. J. Anat. 115: 187–202. common progenitor cell type with mast cells (74), suggesting that 22. Parmley, R. T. 1988. Mammals. In Vertebrate Blood Cells. A. F. Rowley, and basophils and mast cells are derived from a common ancestral cell N. A. Ratcliffe, eds. Cambridge University Press, New York, p. 337–424. type. In fact, ascidian species possess circulating basophil/mast 23. Arinobu, Y., H. Iwasaki, M. F. Gurish, S. Mizuno, H. Shigematsu, H. Ozawa, D. G. Tenen, K. F. Austen, and K. Akashi. 2005. Developmental checkpoints of cell–like cells (59), which are likely to be the common ancestor the basophil/mast cell lineages in adult murine hematopoiesis. Proc. Natl. Acad. of basophil and mast cells (59, 60, 75). Ascidian basophil/mast Sci. USA 102: 18105–18110. cell–like cells have phagocytic activity (62). Thus, Ag presenta- 24. Qi, X., J. Hong, L. Chaves, Y. Zhuang, Y. Chen, D. Wang, J. Chabon, B. Graham, K. Ohmori, Y. Li, and H. Huang. 2013. Antagonistic regulation by the tran- tion might have been attenuated during the process of evolution scription factors C/EBPa and MITF specifies basophil and mast cell fates. Im- from basophil/mast cell ancestors to mammalian basophils, munity 39: 97–110. whereas mast cells maintained their capacity for Ag presentation. 25. Miettinen, M., and J. Lasota. 2005. KIT (CD117): a review on expression in normal and neoplastic tissues, and mutations and their clinicopathologic corre- lation. Appl. Immunohistochem. Mol. Morphol. 13: 205–220. Acknowledgments 26. Sullivan, B. M., and R. M. Locksley. 2009. Basophils: a nonredundant con- tributor to host immunity. Immunity 30: 12–20. We thank Emeritus Prof. Yuzuru Suzuki (University of Tokyo) for inspir- 27. Okayama, Y., and T. Kawakami. 2006. Development, migration, and survival of ing this study, Dr. Masakazu Kondo (National Fisheries University), mast cells. Immunol. Res. 34: 97–115. Prof. Kiyonao Sada, and Dr. Kenji Takeuchi (University of Fukui) for 28. Castells, M. C., D. S. Friend, C. A. Bunnell, X. Hu, M. Kraus, R. T. Osteen, and technical advice, Hideki Uwaoku for kindly providing a healthy fugu K. F. Austen. 1996. The presence of membrane-bound stem cell factor on highly immature nonmetachromatic mast cells in the peripheral blood of a patient with juvenile from the Hatchery Centre of Fukui Prefecture, and Tomohiro aggressive systemic mastocytosis. J. Allergy Clin. Immunol. 98: 831–840. Kawashiro, Toyo Ueda, and Junko Nomura for breeding of fish and main- 29. Ali, H. 2010. Regulation of human mast cell and basophil function by ana- tenance of tanks at the Fukui Prefectural University Research Center for phylatoxins C3a and C5a. Immunol. Lett. 128: 36–45. Marine Bioresources. 30. Suurmond, J., J. N. Stoop, F. Rivellese, A. M. Bakker, T. W. Huizinga, and R. E. Toes. 2014. Activation of human basophils by combined toll-like receptor- and FcεRI-triggering can promote Th2 skewing of naive T helper cells. Eur. J. Immunol. 44: 386–396. Disclosures 31. Sokol, C. L., G. M. Barton, A. G. Farr, and R. Medzhitov. 2008. A mechanism The authors have no financial conflicts of interest. for the initiation of -induced T helper type 2 responses. Nat. Immunol. 9: 310–318. 32. Turner, H., and J. P. Kinet. 1999. Signalling through the high-affinity IgE re- ceptor Fc epsilonRI. Nature 402(6760 Suppl.): B24–B30. References 33. Reese, T. A., H. E. Liang, A. M. Tager, A. D. Luster, N. Van Rooijen, 1. Eberle, J. U., and D. Voehringer. 2016. Role of basophils in protective immunity D. Voehringer, and R. M. Locksley. 2007. Chitin induces accumulation in tissue to parasitic infections. Semin. Immunopathol. 38: 605–613. of innate immune cells associated with allergy. Nature 447: 92–96. 2. Otsuka, A., and K. Kabashima. 2015. Mast cells and basophils in cutaneous 34. Sampson, D., and G. T. Archer. 1967. Release of histamine from human baso- immune responses. Allergy 70: 131–140. phils. Blood 29: 722–736. 10 TELEOST BASOPHILS HAVE A NOVEL DEGRANULATION SYSTEM

35. Stone, K. D., C. Prussin, and D. D. Metcalfe. 2010. IgE, mast cells, basophils, 55. Takaya, K. 1969. The relationship between mast cells and histamine in phy- and . J. Allergy Clin. Immunol. 125(2 Suppl. 2): S73–S80. logeny with special reference to reptiles and birds. Arch. Histol. Jpn. 30: 401– 36. Cromheecke, J. L., K. T. Nguyen, and D. P. Huston. 2014. Emerging role of 420. human basophil biology in health and disease. Curr. Allergy Rep. 14: 56. Sypek, J. P., M. Borysenko, and S. R. Findlay. 1984. Anti-immunoglobulin in- 408. duced histamine release from naturally abundant basophils in the snapping turtle, 37. Yoshimoto, T., K. Yasuda, H. Tanaka, M. Nakahira, Y. Imai, Y. Fujimori, and Chelydra serpentina. Dev. Comp. Immunol. 8: 359–366. K. Nakanishi. 2009. Basophils contribute to T(H)2-IgE responses in vivo via IL- 57. Chieffi Baccari, G., A. de Paulis, L. Di Matteo, M. Gentile, G. Marone, and 4 production and presentation of peptide-MHC class II complexes to CD4+ S. Minucci. 1998. In situ characterization of mast cells in the frog Rana escu- T cells. Nat. Immunol. 10: 706–712. lenta. Cell Tissue Res. 292: 151–162. 38. Sokol, C. L., N. Q. Chu, S. Yu, S. A. Nish, T. M. Laufer, and R. Medzhitov. 2009. 58. Shahid, M., T. Tripathi, F. Sobia, S. Moin, M. Siddiqui, and R. Ali Khan. 2009. Basophils function as -presenting cells for an allergen-induced T helper Histamine, histamine receptors, and their role in immunomodulation: an updated type 2 response. Nat. Immunol. 10: 713–720. systematic review. Open Immunol. J. 2: 9–41. 39. Perrigoue, J. G., S. A. Saenz, M. C. Siracusa, E. J. Allenspach, B. C. Taylor, 59. de Barros, C. M., L. R. Andrade, S. Allodi, C. Viskov, P. A. Mourier, P. R. Giacomin, M. G. Nair, Y. Du, C. Zaph, N. van Rooijen, et al. 2009. MHC M. C. Cavalcante, A. H. Straus, H. K. Takahashi, V. H. Pomin, V. F. Carvalho, class II-dependent basophil-CD4+ T cell interactions promote T(H)2 cytokine- et al. 2007. The Hemolymph of the ascidian Styela plicata (Chordata-Tunicata) dependent immunity. Nat. Immunol. 10: 697–705. contains inside basophil-like cells and a unique sulfated galactoglucan in 40. Otsuka, A., S. Nakajima, M. Kubo, G. Egawa, T. Honda, A. Kitoh, T. Nomura, the plasma. J. Biol. Chem. 282: 1615–1626. S. Hanakawa, C. Sagita Moniaga, B. Kim, et al. 2013. Basophils are required for 60. Cavalcante, M. C., L. R. de Andrade, C. Du Bocage Santos-Pinto, A. H. Straus, the induction of Th2 immunity to and peptide . Nat. Commun. 4: H. K. Takahashi, S. Allodi, and M. S. Pava˜o. 2002. Colocalization of heparin and 1739. histamine in the intracellular granules of test cells from the invertebrate Styela 41. Miyake, K., N. Shiozawa, T. Nagao, S. Yoshikawa, Y. Yamanishi, and plicata (Chordata-Tunicata). J. Struct. Biol. 137: 313–321. H. Karasuyama. 2017. Trogocytosis of peptide-MHC class II complexes from 61. Wong, G. W., L. Zhuo, K. Kimata, B. K. Lam, N. Satoh, and R. L. Stevens. 2014. dendritic cells confers antigen-presenting ability on basophils. Proc. Natl. Acad. Ancient origin of mast cells. Biochem. Biophys. Res. Commun. 451: 314–318. Sci. USA 114: 1111–1116. 62. Garcı´a-Garcı´a, E., N. E. Go´mez-Gonza´lez, J. Meseguer, A. Garcı´a-Ayala, and 42. Ohtani, M., N. Hayashi, K. Hashimoto, T. Nakanishi, and J. M. Dijkstra. 2008. V. Mulero. 2014. Histamine regulates the inflammatory response of the tunicate Comprehensive clarification of two paralogous /13 loci in teleost Styela plicata. Dev. Comp. Immunol. 46: 382–391. Downloaded from fish. Immunogenetics 60: 383–397. 63. Rosenstein, R. K., J. S. Bezbradica, S. Yu, and R. Medzhitov. 2014. Signaling 43. Zhu, L. Y., P. P. Pan, W. Fang, J. Z. Shao, and L. X. Xiang. 2012. Essential role pathways activated by a protease allergen in basophils. Proc. Natl. Acad. Sci. of IL-4 and IL-4Ra interaction in adaptive immunity of zebrafish: insight into USA 111: E4963–E4971. the origin of Th2-like regulatory mechanism in ancient vertebrates. J. Immunol. 64. Seaf, M., M. Ben-Zimra, D. Mankuta, N. Dayan, and F. Levi-Schaffer. 2016. 188: 5571–5584. Papain activates human mast cells to release proinflammatory mediators via its 44. Yamaguchi, T., S. Miyata, F. Katakura, T. Nagasawa, Y. Shibasaki, T. Yabu, enzymatic activity. J. Invest. Dermatol. 136: 1523–1525. U. Fischer, C. Nakayasu, T. Nakanishi, and T. Moritomo. 2016. Recombinant 65. Kim, B. S., K. Wang, M. C. Siracusa, S. A. Saenz, J. R. Brestoff, +

carp IL-4/13B stimulates in vitro proliferation of carp IgM B cells. Fish L. A. Monticelli, M. Noti, E. D. Tait Wojno, T. C. Fung, M. Kubo, and D. Artis. http://www.jimmunol.org/ Shellfish Immunol. 49: 225–229. 2014. Basophils promote responses in inflamed skin. J. 45. Tsujimura, Y., K. Obata, K. Mukai, H. Shindou, M. Yoshida, H. Nishikado, Immunol. 193: 3717–3725. Y. Kawano, Y. Minegishi, T. Shimizu, and H. Karasuyama. 2008. Basophils play 66. Cruikshank, W. W., H. Kornfeld, and D. M. Center. 2000. Interleukin-16. J. a pivotal role in immunoglobulin-G-mediated but not immunoglobulin-E- Leukoc. Biol. 67: 757–766. mediated systemic . Immunity 28: 581–589. 67. Rumsaeng, V., W. W. Cruikshank, B. Foster, C. Prussin, A. S. Kirshenbaum, 46. Da’as, S., E. M. Teh, J. T. Dobson, G. K. Nasrallah, E. R. McBride, H. Wang, T. A. Davis, H. Kornfeld, D. M. Center, and D. D. Metcalfe. 1997. Human mast D. S. Neuberg, J. S. Marshall, T. J. Lin, and J. N. Berman. 2011. Zebrafish mast cells produce the CD4+ T lymphocyte chemoattractant factor, IL-16. J. Immunol. cells possess an FcεRI-like receptor and participate in innate and adaptive im- 159: 2904–2910. mune responses. Dev. Comp. Immunol. 35: 125–134. 68. Abraham, S. N., and R. Malaviya. 1997. Mast cells in infection and immunity. 47. Chen, K., W. Xu, M. Wilson, B. He, N. W. Miller, E. Bengte´n, E. S. Edholm, Infect. Immun. 65: 3501–3508. P. A. Santini, P. Rath, A. Chiu, et al. 2009. enhances immune 69. Suurmond, J., J. van Heemst, J. van Heiningen, A. L. Dorje´e, M. W. Schilham,

surveillance by activating antimicrobial, proinflammatory and B cell-stimulating F. B. van der Beek, T. W. Huizinga, A. J. Schuerwegh, and R. E. Toes. 2013. by guest on September 28, 2021 programs in basophils. Nat. Immunol. 10: 889–898. Communication between human mast cells and CD4+ T cells through antigen- 48. Edholm, E. S., E. Bengte´n, J. L. Stafford, M. Sahoo, E. B. Taylor, N. W. Miller, dependent interactions. Eur. J. Immunol. 43: 1758–1768. and M. Wilson. 2010. Identification of two IgD+ B cell populations in channel 70. Ellis, J. S., F. B. Guloglu, D. M. Tartar, C. M. Hoeman, C. L. Haymaker, catfish, Ictalurus punctatus. J. Immunol. 185: 4082–4094. J. A. Cascio, X. Wan, M. Dhakal, A. VanMorlan, S. H. Yahng, and H. Zaghouani. 49. Edholm, E. S., E. Bengten, and M. Wilson. 2011. Insights into the function of 2010. APCs expressing high levels of programmed death ligand 2 sustain the IgD. Dev. Comp. Immunol. 35: 1309–1316. development of CD4 T cell memory. J. Immunol. 185: 3149–3157. 50. Lantz, C. S., M. Yamaguchi, H. C. Oettgen, I. M. Katona, I. Miyajima, 71. Ding, Q., L. Lu, B. Wang, Y. Zhou, Y. Jiang, X. Zhou, L. Xin, Z. Jiao, and J. P. Kinet, and S. J. Galli. 1997. IgE regulates mouse basophil Fc epsilon RI K. Y. Chou. 2006. B7H1-Ig fusion protein activates the CD4+ IFN-gamma re- expression in vivo. J. Immunol. 158: 2517–2521. ceptor+ type 1 T regulatory subset through IFN-gamma-secreting Th1 cells. J. 51. MacGlashan, D., Jr., H. Z. Xia, L. B. Schwartz, and J. Gong. 2001. IgE-regulated Immunol. 177: 3606–3614. loss, not IgE-regulated synthesis, controls expression of FcepsilonRI in human 72. Laidlaw, B. J., W. Cui, R. A. Amezquita, S. M. Gray, T. Guan, Y. Lu, basophils. J. Leukoc. Biol. 70: 207–218. Y. Kobayashi, R. A. Flavell, S. H. Kleinstein, J. Craft, and S. M. Kaech. 2015. 52. Borkowski, T. A., M. H. Jouvin, S. Y. Lin, and J. P. Kinet. 2001. Minimal re- Production of IL-10 by CD4+ regulatory T cells during the resolution of infection quirements for IgE-mediated regulation of surface Fc epsilon RI. J. Immunol. promotes the maturation of memory CD8+ T cells. Nat. Immunol. 16: 871–879. 167: 1290–1296. 73. Denzel, A., U. A. Maus, M. Rodriguez Gomez, C. Moll, M. Niedermeier, 53. Reite, O. B. 1965. A phylogenetical approach to the functional significance of C. Winter, R. Maus, S. Hollingshead, D. E. Briles, L. A. Kunz-Schughart, et al. tissue mast cell histamine. Nature 206: 1334–1336. 2008. Basophils enhance immunological memory responses. Nat. Immunol. 9: 54. Mulero, I., M. P. Sepulcre, J. Meseguer, A. Garcı´a-Ayala, and V. Mulero. 2007. 733–742. Histamine is stored in mast cells of most evolutionarily advanced fish and reg- 74. Galli, S. J. 2000. Mast cells and basophils. Curr. Opin. Hematol. 7: 32–39. ulates the fish inflammatory response. Proc. Natl. Acad. Sci. USA 104: 19434– 75. Crivellato, E., and D. Ribatti. 2010. The mast cell: an evolutionary perspective. 19439. Biol. Rev. Camb. Philos. Soc. 85: 347–360. Supplemental Table I. Primer sets used for RT-PCR and quantitative RT-PCR in this article.

RT-PCR Quantitative RT-PCR Gene Sequence, 5’-3’ Gene Sequence, 5’-3’ actb-F AGCCAACAGGGAGAAGATGACCC ef1a-F AGCAAGAACGACCCACCAA actb-R GAGCGT AGCCCTCGTAGATGGG ef1a-R TGCAGGCAATGTGAGCTGT cebpa-F ACAGTGCACGACAGGACAAG tcra-F AGCGCATGTCTGGCCACAGGTTTCA cebpa-R GGCTGAAGGTGCATGGTAGT tcra-R GACTGATACGCAGACGAAGAGTCATCAGG kit-F CACTGATGTCCGAGCTGAAA tcrd-F ACGGACCAGGAGTGGACA kit-R TCATGTAGCCGTTCAAGCTG tcrd-R GTGAGGACCACCCTCAACAT cd83-F ATGTCTCCTTGGTTCTTCACG cd4-F GGCTGCGGGGAAAGTTTAGA cd83-R TTACACACAGATGTGCTTAGTTTTGGGTT cd4-R AACAGGAAGAACGTGGTCGA mpo-F TTACCAGCGTCTGCAACAAC cd8a-F CGAGGAGAAAATGGAGCAAAAGTGGAGG mpo-R TTATCCTCGGAGCCGTACAC cd8a-R GCACCATTTCGACATAACTTCTCCGACGG lcp1-F GGGTCAACGAGACTCTCAGG prf1-F AGCTGGCGGAGTATTCCATG lcp1-R CACCGTCATCACCATCTTTG prf1-R GTTGCTCAGCTTGGGAGAGT ighm-F CTGGAGCCCCAATCTAAAACCTACG ighm-F TGTGCATGCGGAAGAAGACA ighm-R CTAGACAGCGAATGACATCGTACGGAC ighm-R GTGGTGGCAATGCTGAACAG cd3e-F CCAACCGATTAGAGCGAATCAGAGGC ight-F TGTTCTGTGTCTCACGAGGG cd3e-R CCTGATTCCTTCCCGGATCCAGC ight-R TGGCTTCACTGGACCTCTCA tcra-F AGCGCATGTCTGGCCACAGGTTTCA tbx21-F AGGTACCAAGGTGGCAAGTG tcra-R GACTGATACGCAGACGAAGAGTCATCAGG tbx21-R GGTGCATGTACATCCTGTTCC prf1-F CCGAAATGAAGCACTGCAAG gata3-F CCTGCGGCCTCTACTTCAAA prf1-R TCTTCCACAGGCCTCTCTCC gata3-R AAGTCGTCCATGCTGTCCTG mhc2a-F CCCGTTAACGTCTCCTGGACCAGGAACG foxp3-F GCCGCTCACGCTCTCTAC mhc2a-R CAGGCTGCTGCACCTCCACAGTCCACAT foxp3-R TGCTTGATGAAGTGGTTGAAA il4/13a-F TCACTGGGAACCTCACCAC ccr7-F TTTCAGCTGCAGCACAACAC il4/13a-R GTACTGCAGGACGTCTTGG ccr7-R CTGGCAGGAGCATCTGAAGG il4/13b-F CATCATTCAAAGTCTGGCACTGA il4/13b-R ATCGGTCAAATTCAGGGACTTCTC hdc-F ATGTGATGGACTGGTTGTGTAAAG hdc-R CCTCCTCTATGGCTTGATTCAGAG il16-F CGTGAAATAGCGGATTTGGT il16-R ACCCACAGTGGAACAGGAAG A D Lymphocyte Monocyte 10 8 15 b b cells 8

6 6 10 6

4 ng/10 4 a, b 5 a

2 a, b a a Migration Migration index 2 Migration index a 0 0

Histamine Histamine ND 0 Ba Pa BaP BaIP Ba Pa BaP BaIP Basophilsbasophils PBLksPBLs Other 15 b 6 b B Basophils PBLks

index 10 4

5 a 2 a a, b

a a Migration Migration

a Migration index 0 0 Ba- Pa- BaP+ BaIP+ Ba- Pa- BaP+ BaIP+ Basophil - + + - - + + - Papain - - - + - - - + HI-papain

E C Individual no. 100100% % M 1 2 3 NTC

880%0 % hdc Lymphocyte Monocyte 6060% % Neutrophil actb 440%0 % Other

2020% %

00% % Ba- Pa- BaP+ BaIP+ Basophil - + + - Papain - - - + HI-papain

Supplemental Figure 1. The red granules of fugu basophils stored histamine, and degranulation of the granules activated various leukocytes migration. (A) Histamine was detected in the basophil lysate. Histamine concentration in the cell lysate was measured with EIA. Error bars show the mean ± SEM of results from three individuals. ND, not detected. (B) O-phthalaldehyde stained fugu basophils, but not PBLks. Scale bar, 10 μm. (C) Histidine decarboxylase was expressed in fugu basophils. M is the molecular size marker. NTC indicates the no-template control. (D) Chemotaxis of PBLks to the culture supernatant of papain stimulated fugu basophils. PBLks were classified into four populations by morphology, and counted. Values were normalized to those of the non-stimulated control. Error bars show the mean ± SEM of results from four individuals. P < 0.05. (E) Each population ratio in migrated leukocytes. The data are representative results from four individuals. Normal mouse serum Anti-fugu IgM mAb

-IgM

+IgM

Supplemental Figure 2. Fugu basophils degranulated by the crosslinking of IgM with an anti-fugu IgM mAb. Arrows indicate degranulated basophils. Scale bar, 10 μm. The data are representative results from three individuals.